Algorithm Could Help Enable Next-Generation Deep Brain Stimulation

New means to stimulate brains.  As I read it, this may require quite a bit of regulation. 

Algorithm Could Help Enable Next-Generation Deep Brain Stimulation Devices

By News from Brown University, June 8, 2021

Brown University bioengineers have developed a new algorithm that could clear a path to more adaptive deep brain stimulation (DBS) technology.  The algorithm helps DBS systems more easily detect brain signals while concurrently delivering stimulation, by identifying and eliminating electrical artifacts.

The Brown team was able to use the algorithm to stitch fragments of low-resolution data into a high-resolution picture of an artifact waveform, which outperformed other approaches in distinguishing brain signals from artifacts in laboratory experiments and computer simulations.  Brown's Nicole Provenza said this differentiation holds up even when the signal of interest is very similar to simulation artifacts.

The researchers also said the algorithm is computationally inexpensive, which suggests the possibility of real-time artifact-filtering, and simultaneous recording and stimulation.

  (full article) 

Spy Agencies look to Science

When there are no direct solutions, look to emerging tech.  But do they have the same goals?   Same limitations? 

Spy Agencies Turn to Scientists as They Wrestle With Mysteries

By The New York Times via CACM   July 9, 2021

The nation's intelligence agencies are looking for ways to increase their expertise in a range of scientific disciplines as they struggle to answer unexplained questions — about the origins of the coronavirus pandemic, unidentified phenomenon observed by Navy pilots, and mysterious health ailments affecting spies and diplomats around the world.

Traditional spycraft has failed to make significant progress on those high-profile inquiries, and many officials have grown convinced that they require a better marriage of intelligence gathering and scientific examination.

Intelligence officials in the Biden administration came into office pledging to work on areas traditionally dominated by science, like studying the national security implications of climate change and future pandemics. But as the other issues have cropped up, the spy agencies have had to confront questions that are as much scientific mysteries as they are challenges of traditional intelligence collection.

The White House has given the intelligence community until later this summer to report the results of a deep dive into the origins of the coronavirus, including an examination of the theory that it was accidentally leaked from a Chinese lab studying the virus as well as the prevailing view that it was transmitted from animals to humans outside a lab.

In the NYT

Bio Neurons versus Computational

When we first learned of the use of the patterns of brain neurons to potentially use as reasoning devices, we took a course from actual neuroscientists.  We very quickly learned that human bio neurons were very much more complex that neurons in our feeble  'neural networks'.   Always been intrigued by the concept, so how could they be more useful 'reasoners"? Tried to augment  our nets with this new complexity.  Here is another case why they are very different.  We can ask ourselves, can we use these aspects of neurons to improve reasoning?  Note the indication that timing matters.  What embedded information  could make us learn faster?   Still unclear, 

Neurons Unexpectedly Encode Information in the Timing of Their Firing

Elena Renken  Quanta Mag     Contributing Writer

A temporal pattern of activity observed in human brains for the first time may explain how we can learn so quickly.

For decades, neuroscientists have treated the brain somewhat like a Geiger counter: The rate at which neurons fire is taken as a measure of activity, just as a Geiger counter’s click rate indicates the strength of radiation. But new research suggests the brain may be more like a musical instrument. When you play the piano, how often you hit the keys matters, but the precise timing of the notes is also essential to the melody.

“It’s really important not just how many [neuron activations] occur, but when exactly they occur,” said Joshua Jacobs, a neuroscientist and biomedical engineer at Columbia University who reported new evidence for this claim last month in Cell.  ... ' 

Evolving Bio Intelligence

Can this kind of generalized AI take us towards more generalized AI?   A new shift in direction? 

Potcast:  Bio Eats World: Evolving Embodied Intelligence    

by Li Fei-Fei, Surya Ganguli, Vijay Pande, and Lauren Richardson  By Andreessen, A16z

On today’s episode, we are making the full arc from the theoretical and borderline philosophical to the applied. Let’s start with the theory: embodied intelligence posits that the body, or the physical form, plays an active and significant role in shaping an agent’s mind and cognitive capacities. For example, human intelligence is not just the function of our brain, but a combination of our brain, our body, and the environment in which we exist. But when it comes to designing artificial intelligence (AI), a physical form and an environment are typically not part of the equation. It’s a disembodied cognition. Our guests, Li Fei-Fei and Surya Ganguli of the Stanford Institute for Human-Centered AI,   set out to develop what they call an “evolutionary playground” to explore the development of embodied intelligence in AI and its connection with the environment and with learning using in silico experiments. They discuss with a16z general partner Vijay Pande and host Lauren Richardson how they created a suite of virtual environments in which agents evolve through a process that mimics aspects of Darwinian evolution. These agents, called the unimal, or universal animal, start off as a central node, and with each generation can add or subtract limbs and change various properties of their physical forms, like how flexible their joints are. Just like in real evolution, different forms arose based on the particularities of the environment, but what is really exciting is what Fei-Fei, Surya, and colleagues discovered about the intelligence encoded in some of these forms, such as an increased ability to learn a novel task. Which brings us to the applied section of our discussion. These results provide new insights for how we think about designing robots capable of performing unique tasks, and for understanding the possible limitations of disembodied AI models, like GTP-3.

The results are described in the pre-print “Embodied Intelligence via Learning and Evolution” posted on arXiv.org.

China Rise in Smart-Cities

Are we too regulated, too cautious of uses of automation?  Interesting smart city stats. 

A Global Smart-City Competition Highlights China's Rise in AI,

By Wired, July 9, 2021  in ACM/Wired

Four years ago, organizers created the international AI City Challenge to spur the development of artificial intelligence for real-world scenarios like counting cars traveling through intersections or spotting accidents on freeways.

In the first years, teams representing American companies or universities took top spots in the competition. Last year, Chinese companies won three out of four competitions.

Last week, Chinese tech giants Alibaba and Baidu swept the AI City Challenge, beating competitors from nearly 40 nations. Chinese companies or universities took first and second place in all five categories. TikTok creator ByteDance took second place in a competition to identify car accidents or stalled vehicles from freeway videofeeds.

The results reflect years of investment by the Chinese government in smart cities. Hundreds of Chinese cities have pilot programs, and by some estimates, China has half of the world's smart cities. The spread of edge computing, cameras, and sensors using 5G wireless connections is expected to accelerate use of smart-city and surveillance technology.

From Wired

Document Data Model

Good overview .... and reminder of usefulness.

WHITE PAPER

Why the Document Data Model

Built around JSON-like documents, document databases are both natural and flexible for developers to work with. They promise higher developer productivity, and faster evolution with application needs. As a class of non-relational, sometimes called NoSQL database, the document data model has become the most popular alternative to tabular, relational databases.

In this guide, we’ll discuss the fundamental design concepts underpinning the document data model, how they help developers innovate faster, and use cases where you can put document databases to work.

Download the Why Documents Guide to learn more.  ... 

Digital Twins

Useful piece on Digital Twins and their definition and use.   Adds to my current examination and use cases.

The Multiple Faces of Digital Twins   By Alex Woodie  in Datanami

Digital twins are emerging as a hot technology, particularly among manufacturers and companies involved with the Industrial Internet of Things. Depending on the use cases, though, customers may opt for one type of digital twin over another.

To a certain extent, every digital twin is a unique creation. The ability to create a digitized copy of an actual physical asset, such as a wind turbine or a locomotive, and measure how that model responds and reacts to different inputs is the fundamental breakthrough that is driving adoption of digital twin technologies.

But there are a few broad categories of digital twins, and companies that are considering adopting a digital twin would do well to explore how their use cases match up to these types.

According to Philipp Wallner, the industry manager for MathWorks, there are two general types of digital twins: physics-based twins and data-based twins.

The physics-based digital twin is a detailed reproduction of a well-understood piece of machinery that behaves in a predictable manner, such as a robot or a piece of equipment on a manufacturing line. In some cases, these physics-based digital twins are created by importing CAD files into a simulation platform.

Data-based twins, on the other hand, are models that machineries or processes that are not as well understood and have very complex interactions, such as a compressor. While the physics driving these machines are generally understood, the high number of variables involved, including the shapes of vessels, precludes basing the model directly on physics. However, users do have a large amount of data describing the machine’s behavior, which becomes the basis for the model.

MathWorks helps its customers build both types of digital twins. One of its customers, Krones, developed a digital twin in MathWorks’ software based on CAD drawings of its bottle-handling robot.

“What we offer is a CAD import functionality,” Wallner says. “So you take these CAD models, and you import them. And then you already have a pretty good basis for your model. You still have to add additional parameters and some of the functionality that is not captured in the CAD model.”  ... ' 

McKinsey: Trends in Tech

Well done overview:

The top trends in tech

Which technologies have the most momentum in an accelerating world? We identified the trends that matter most

The top trends in tech

Which technologies have the most momentum in an accelerating world? We identified the trends that matter most.

McKinsey tech trends index

Applied 

AI

Next-generation computing

Trust architecture

Distributed infrastructure

Future of connectivity

Future of programming

Next-level process automation and virtualization

Bio Revolution

Future of clean  technologies

Nanomaterials

As all things digital continue to accelerate, which technology trends matter most for companies and executives? To answer that question, we developed a unique methodology to identify the ten trends most relevant to competitive advantage and technology investments. 

These trends may not represent the coolest, most bleeding-edge technologies. But they’re the ones drawing the most venture money, producing the most patent filings, and generating the biggest implications for how and where to compete and the capabilities you need to accelerate performance. 

Unifying and underlying them all is the combinatorial effect of massively faster computation propelling new convergences between technologies; startling breakthroughs in health and materials sciences; an array of new product and service functionalities; and a strong foundation for the reinvention of companies, markets, industries, and sectors   .... " 

Data Security for Payment Processing

A space I worked in for a while.  There is a lot of security that is required.  

Data Security Rules Instituted for U.S. Payment Processing System  By ZDNet  July 9, 2021

New data security rules governing the payment system that facilitates direct deposits and direct payments for nearly all U.S. bank and credit union accounts are now in effect.

The National Automated Clearinghouse Association (NACHA) stipulates that an account number used for any Automated Clearinghouse (ACH) payment must be rendered indecipherable while stored electronically.

This mandate is applicable to any facility where account numbers related to ACH entries are stored.

NACHA has instructed ACH originators and third parties that process over 6 million ACH transactions annually to render deposit account data unreadable when stored electronically, recommending measures that include encryption, truncation, tokenization, and destruction.

The regulator said access controls like passwords are unacceptable, but disk encryption is permitted, provided additional and prescribed physical safeguards are implemented.

in ZDNet  ... 

Need of Advances for Electric Trucks

 On the use of electric trucks, and what they will need to make them broadly viable.

Electric trucks can go the (short) distance

To electrify the entire industry, battery technology and charging infrastructure would have to improve.  by Casey Crownhart

While some trucks cover over a thousand miles in a day, others operate at short range. These vehicles are more feasible for electrification in the near term, according to Brennan Borlaug, a researcher at the US National Renewable Energy Laboratory and the lead author of a recent Nature Energy study on electric-truck charging infrastructure.        ... " 

Building a Better Wallet

 Digital wallets in the past have been insecure on some systems, and they are the primary user interface into cryptocurrency. 

Square is building a hardware crypto wallet and service with the goal of making bitcoin 'more mainstream'

By Tyler Sonnemaker in BusinessInsider

Square plans to build a hardware cryptocurrency wallet and software service, company execs said Thursday.

Square's goal is to "make bitcoin custody more mainstream," said hardware lead Jesse Dorogusker.

CEO Jack Dorsey in June outlined the company's initial thinking about the product.    ... '

Identification Improvements Using Blurred Fingerprints

Fingerprint clear up analysis.    An example of the use of Generative Adversarial Networks.  But note the comment on auditing the results in court.

AI Clears Up Images of Fingerprints to Help with Identification

New Scientist, Matthew Sparkes, June 28, 2021

West Virginia University researchers have trained an artificial intelligence (AI) model to clean up distorted images of fingerprints from crime scenes to improve identification. The researchers developed a generative adversarial network by creating blurred versions of 15,860 clean fingerprint images from 250 subjects. They trained the AI using nearly 14,000 of these pairs of images; when they tested its performance on the remainder, they found the model to be 96% accurate at the lower end of the range of blurring intensity, and 86% at the higher end. Forensic Equity's David Goodwin said the use of neural networks to manipulate images would have trouble standing up in court because they cannot be audited like human-generated code, and the inner workings of these models are unknown.   ... '

Challenge for Learning from Human Feedback using Minecraft

Berkeley Bair challenge competition here using a common gaming environment.   Been a long time since I looked at Minecraft.  Short extract of the idea below, more complete look at the link.   Seems a novel look at a broader look at contextual learning. 

 BASALT: A Benchmark for  Learning from Human Feedback   by Rohin Shah    Jul 8, 2021

TL;DR: We are launching a NeurIPS competition and benchmark called BASALT: a set of Minecraft environments and a human evaluation protocol that we hope will stimulate research and investigation into solving tasks with no pre-specified reward function, where the goal of an agent must be communicated through demonstrations, preferences, or some other form of human feedback. Sign up to participate in the competition!

Motivation

Deep reinforcement learning takes a reward function as input and learns to maximize the expected total reward. An obvious question is: where did this reward come from? How do we know it captures what we want? Indeed, it often doesn’t capture what we want, with many recent examples showing that the provided specification often leads the agent to behave in an unintended way.

Our existing algorithms have a problem: they implicitly assume access to a perfect specification, as though one has been handed down by God. Of course, in reality, tasks don’t come pre-packaged with rewards; those rewards come from imperfect human reward designers.

For example, consider the task of summarizing articles. Should the agent focus more on the key claims, or on the supporting evidence? Should it always use a dry, analytic tone, or should it copy the tone of the source material? If the article contains toxic content, should the agent summarize it faithfully, mention that toxic content exists but not summarize it, or ignore it completely? How should the agent deal with claims that it knows or suspects to be false? A human designer likely won’t be able to capture all of these considerations in a reward function on their first try, and, even if they did manage to have a complete set of considerations in mind, it might be quite difficult to translate these conceptual preferences into a reward function the environment can directly calculate.  ...................

Conclusion

We hope that BASALT will be used by anyone who aims to learn from human feedback, whether they are working on imitation learning, learning from comparisons, or some other method. It mitigates many of the issues with the standard benchmarks used in the field. The current baseline has lots of obvious flaws, which we hope the research community will soon fix.

Note that, so far, we have worked on the competition version of BASALT. We aim to release the benchmark version shortly. You can get started now, by simply installing MineRL from pip and loading up the BASALT environments. The code to run your own human evaluations will be added in the benchmark release.

If you would like to use BASALT in the very near future and would like beta access to the evaluation code, please email the lead organizer, Rohin Shah, at rohinmshah@berkeley.edu.

This post is based on the paper “The MineRL BASALT Competition on Learning from Human Feedback”, accepted at the NeurIPS 2021 Competition Track. Sign up to participate in the competition!   

Algorithmic Impact Assessments

Welcome  to Algorithmic Impact Assessments.  I should note that I have done optimization and statistical models for decades, and this has very, very rarely come up.   And then only when included in a risk analysis that indicated potential errors being made.   Will all business process be under detailed scrutiny?    Article linked to below is interesting.

Assembling Accountability, from the Ground Up     in  Point.datasociety.net

Algorithmic impact assessments should leverage diverse expertise & complex histories

By Emanuel Moss, Ranjit Singh, Elizabeth Anne Watkins, and Jacob Metcalf

By now, it is a tired trope: Sen. Orrin Hatch asking Mark Zuckerberg how Facebook makes money. Zuckerberg replying with a wry “Senator, we run ads.” Another congressperson grilling Sundar Pichai, CEO of Google, about the Apple iPhone, made by Apple.

Lawmakers, it is said, don’t understand technology well enough to regulate it. They are too old. They are out of touch. They have disinvested from staff and other experts that could help them understand it. And while all those criticisms may be true, why should we expect our lawmakers to become individual experts on every challenge facing society? There are advocacy groups, community activists, forensic technologists, thoughtful developers, and critical scholars who have devoted their careers to building expertise on these issues. Given the burgeoning influence of algorithmic systems over social affairs, and an increasing awareness of the harmful impacts of these powerful systems, we are at a moment in which complex sociotechnical systems require robust, adaptable regulation — and legislatures and regulatory bodies are drafting new rules.

Our new report Assembling Accountability  demonstrates a pressing need to establish algorithmic impact assessment practices from the ground up, which requires cultivating and synthesizing a broad consensus of expertise from industry, scholars, and public interest advocates, including people from affected communities.  ... ' 

IBM Using Blockchain for Validation

A trust/validation type of application for blockchain.     In Coindesk

IBM, Heifer International to Assist Honduras Farmers Access Global Markets Using Blockchain

IBM's Food Trust network will help coffee and cocoa farmers and buyers verify information on their supply chains. ... '

Will AI Rewrite Coding?

Inclined to think so. Increase efficiency, prevent errors and better ensure secure practices. Below intro describes the current progress and direction.  Starting with assistant approaches and code checking.  Note the mention of a number of companies and projects involved.   More at the link.

Will AI Rewrite Coding?   By Samuel Greengard, Commissioned by CACM Staff,   July 6, 2021

Computer code now touches almost every aspect of our lives. Worldwide, 27 million developers churn out billions of lines of code every day. Yet, despite an abundance of open source libraries and increasingly sophisticated development tools, the task is time-consuming and prone to errors.

As a result, researchers are studying ways to introduce artificial intelligence (AI) into coding processes. While much of the effort centers on automating coding tasks, spotting bugs, fixing vulnerabilities, and producing more elegant code, there's also an emerging effort to tap AI to write code based on short text descriptions of what the code should do.

"There's interest in improving current coding practices and generating code through machine learning and AI models," says Brendan Dolan-Gavitt, an assistant professor of computer science and engineering at New York University (NYU) Tandon School.

Adds Furkan Bektes, founder and CEO of SourceAI, which has developed a tool to write code based on short natural language input, "The use of AI will allow developers to code faster, and allow non-developers to pursue their ideas."

Code of Conduct

The complexities of today's coding processes are not lost on anyone. A Boeing 787 Dreamliner has approximately 20 million lines of code. Major software programs and games have upwards of 50 million lines of code. Somewhere between functionality and chaos lies the real-world task of producing code rapidly and as bug-free as possible.

AI is taking direct aim at the challenge. "Ideally, AI could intervene, examine patterns and provide feedback about coding errors," says Shashank Srikant, a Ph.D. student in the Department of Electrical Engineering and Computer Science at the Massachusetts Institute of Technology (MIT). "This could help coders avoid traps that others have fallen into."

Google, Microsoft, and others have already begun experimenting with AI for assisted coding. For instance, in 2018, Microsoft introduced AI-assisted coding for Java, Python, C++, and other languages through Visual Studio IntelliCode. It offers developers relevant coding suggestions based on thousands of the most popular open source projects at GitHub. Through machine learning, it analyzes common usage patterns and practices, and delivers suggestions tailored to a specific project.

However, in May 2020, the field took a giant leap forward. OpenAI introduced a next-generation AI-based neural network and programming model called GPT-3, which is already used to build apps—including buttons, colors and input fields—using AI. When researchers and coding experts began testing the language, they realized that it could also write its own code.

Bektes, among the first to gain access to the platform, fed high-quality code samples into GPT-3 and built an application that generates code in any programming language—using input in English and most other major languages. For example, a user might say, "Calculate factorial of number given by user," and Source AI spits out the code.

Other tools incorporating AI are also popping up. For instance, Tabnine can autofill lines and functions as developers type.

Machine-generated coding could be a game changer, although it's unlikely to supplant the need for developers anytime soon. "It will open new horizons," Bektes explains. "There are many non-developers who have ideas but don't know how to code, and there are also developers who are experts in one language but not in others. AI can help them learn to code in other languages."      ...... ' 

Automotive Cybersecurity

Vehicles in particular will require specialized cybersecurity as their autonomy increases.

Keeping Control of the Wheel   By David Geer   in ACM

The rising need for cybersecurity will trigger investments over the next few years. We expect to see the market grow from US$4.9 billion in 2020 to US$9.7 billion in 2030, with software business representing half of the market by 2030," according to "Cybersecurity in automotive: Mastering the challenge," a 2020 market study by global management consulting firm McKinsey & Company.

The study "Automotive Cybersecurity Market: the Development of Autonomous Cars and Other Notable Growth Drivers," by market intelligence firm Infinity Research, identifies the market forces advancing automotive cybersecurity as including:

The development of autonomous vehicles with wireless connections.

The increasing number of in-vehicle electronic control units and their wireless connections.

Regulatory mandates and standards targeting the cyber-safety of vehicles and data .

Nobody wants criminal hackers in the driver's seat. "Much of the motivation to implement enhanced security systems stems from advances in in-vehicle capabilities. Progress in these internal capabilities includes Advanced Driver Assistance Systems (ADAS). These systems necessitate heightened computer control over sensitive actuators (drive by wire, including steer by wire, throttle by wire, and brake by wire)," says Josh Siegel, assistant professor of computer science and engineering at Michigan State University (MSU).

According to the 2021 HSB Cyber Car Tech Survey by cyber risk insurer HSB Group, more than a third of U.S. consumers say they are concerned about the cybersecurity of connected cars. Another third say they fear a computer virus, hacking incident, or other cyberattack that could damage or destroy their vehicle's data, software, or operating systems.

To MSU's Siegel, growing hacker expertise suggests automotive cyberattacks are unleashed by criminal hackers with malicious intent, and not just discovered by researchers to bring vulnerabilities to light. There have been targeted hacks turning vehicles into espionage devices at military bases and disabling engines, so the individual has to take other transportation to work, says Siegel. "I assume that nation-states or well-resourced entities are executing these attacks," says Siegel.

The 2021 Global Automotive Cybersecurity Report by connected vehicle cybersecurity provider Upstream Security, found that malicious blackhat hackers last year carried out 55% of automotive cyberhacks to disrupt business, steal property, and demand ransom. Whitehat hackers and researchers, including those participating in automotive bug bounty programs, performed 38.6% of hacks, the report says. Bug Bounty programs pay white hat hackers a reward or "bounty" for finding critical vulnerabilities in an organization's software.  ... '

Victims Unaware

Surprising given todays world. 

Data Breaches: Most Victims Unaware When Shown Evidence of Multiple Compromised Accounts

By University of Michigan News, June 23, 2021

Most of the 413 people asked by researchers about breaches that involved their own personal information for a new study were unaware of them.

Most data breach victims do not realize their personal data has been exposed in five breaches on average, according to a study by an international team of researchers.

Investigators at the University of Michigan, George Washington University (GW), and Germany's Karlsruhe Institute of Technology showed 413 participants facts from up to three breaches involving their personal information, and 74% said they were unaware of the breaches.

Most victims blamed their personal habits for the problem—like using the same password for multiple accounts—while just 14% blamed it on external factors.

GW's Adam Aviv said, "The fault for breaches almost always lies with insufficient security practices by the affected company, not by the victims of the breach."

From University of Michigan News

Humanoid Pepper Robot Pauses

 Saw this demonstrated shortly after its creation.   An example of a humanoid robot we thought of using in the retail aisle.  Saw it demonstrated in Japan. Mostly being used for eldercare applications.   Appropriately as part of a smart home.   Appropriately friendly.  Will it be back?

Humanoid Robot Pepper Pauses Production

ERIC HAL SCHWARTZ in Voicebot.ai

The humanoid robot named Pepper has been semi-retired by SoftBank, at least for now. The Japanese corporation has decided to pause production until they see a need for more, even as interest in virtual humans and robot assistants has mushroomed over the last couple of years.

PEPPER POP

Softbank began manufacturing Pepper back in 2014, but only ever made 27,000, according to Reuters. The $1,790 robot requires a $360 a month subscription to function, which might have been considered too high a price for the educational and research facilities most interested in Pepper and its emotion-reading AI. The four-foot 62-pound robot did make plenty of other appearances for other purposes, however. Pepper promoted AI in education by testifying to a British parliamentary select committee and served as a greeter at the HSBC bank in New York and both the Munich and Montreal airports. Several Pepper robots even substituted as cheerleaders for a baseball game last summer during the height of the COVID-19 pandemic. Though Pepper appeared to be autonomous, the robot was often controlled remotely, implying that the AI was smarter and more functional than in reality.

ROBOTIC CARE

The pause on Pepper and its limitations haven’t dampened interest in the idea of mobile, AI-enabled helpers. Healthcare in particular has a lot of opportunities for combinations of robotic bodies and advanced virtual assistants. They vary in detail, but all tend to offer communications and caregiving in some form. Their bodies include the scooter mixed with a tablet computer named temi, which follows owners about the home, and the cat-faced Mylo from CR Robotics, designed to be a friendly, comforting image for people with Alzheimer’s to interact with on a daily basis. For a more explicitly humanoid robot, Singularity Studio and Hanson Robotics created Grace, which looks like the head and upper torso of a woman and can converse with users about a range of subjects, particularly physical and mental health. ... '

For Those Seeking Chatbot ROI

Good general piece on approaches and the ROI of Chatbots.  Could be a good place to start.  From the Alexa Blog, but  appears to be regularly applicable ... 

 Chatbot ROI: What You Need to Know and Expect

Chatbots have become an increasing fixture for businesses in recent years. From offering in-depth customer information to valuable marketing content, chatbots are serving a number of functions on company websites and other owned channels.

But many businesses still want to know if the juice is really worth the squeeze when it comes to chatbots. Is the chatbot ROI they receive going to be worth it?

The answer isn’t a simple yes or no. Chatbots can take on many different forms, levels of AI sophistication, and required maintenance.

To get an accurate assessment of chatbot ROI as a whole, start by examining key use cases and pricing factors. To get the most value out of your chatbot within your budget, you need to know why you’re investing in one and the typical costs for this technology. From there, we’ll help you calculate your chatbot ROI.

Determine Your Need for a Chatbot First

To make your chatbot a more successful investment, it’s important that you have a forward-thinking approach. Have a clear understanding of why you need a chatbot to determine how you should invest in the technology.

Assess how quickly you’ll need the chatbot based on your use case and business’s growth. You might conclude that you don’t need to invest in a chatbot right away, but it’s something you’ll need in six months. Allocate some time to examine different chatbot options to avoid rushed, last-minute research when your business needs the tool. ....'   (See the link above for entire article) ...

Automated Vertical Farming

General description without details, but interesting application.   Was involved with some farming applications and it seems some of these can be readily automated with such systems.  Note posts regarding 'agriculture' here.

Robots Take Vertical Farming to New Heights    in ACM By Governing, July 1, 2021

Fifth Season, a vertical farm in Braddock, PA, uses robots to grow greens indoors, from seed to harvest.

Co-founder Austin Webb said, "What we have built is the industry first, and industry only, end-to-end automated platform."Plastic trays with unique IDs are stacked throughout the 60,000-square-foot facility, and every plant can be traced from any point in the growing process.

Proprietary software directs the robots to stack and remove trays of plants.   LEDs are used to replicate sunlight, and sensors are used to monitor the nutrient mix, carbon dioxide levels, and light spectrum.

Compared to conventional farming, Fifth Season uses up to 95% less water, 98% less land, and no herbicides or pesticides.

Moreover, a half-acre indoors is equivalent to almost 100 acres of farmland in terms of production.

From Governing  ... 

NVIDIA Pretrained Models

Elements of Pretrained models.  Have not used this approach before, worth understanding its application. 

Jun 24, 2021

Fast-Track Production AI with Pretrained Models and Transfer Learning Toolkit 3.0  By Joanne Chang

Tags: Computer Vision & Machine Vision, DeepStream, featured, Healthcare & Life Sciences, Jarvis, Manufacturing, Metropolis, News, NLP / Conversational AI, Pretrained Models, Retail/Etail, Robotics, TLT, Triton

Today, NVIDIA announced new pretrained models and general availability of Transfer Learning Toolkit (TLT) 3.0, a core component of NVIDIA’s Train, Adapt, and Optimize (TAO) platform guided workflow for creating AI. The new release includes a variety of highly accurate and performant pretrained models in computer vision and conversational AI, as well as a set of powerful productivity features that boost AI development by up to 10x. 

As enterprises race to bring AI-enabled solutions to market, your competitiveness relies on access to the best development tools. The development journey to deploy custom, high-accuracy, and performant AI models in production can be treacherous for many engineering and research teams attempting to train with open-source models for AI product creation. NVIDIA offers high-quality, pretrained models and TLT to help reduce costs with large-scale data collection and labeling. It also eliminates the burden of training AI/ML models from scratch. New entrants to the computer vision and speech-enabled service market can now deploy production-class AI without a massive AI development team. 

Highlights of the new release include:

A pose-estimation model that supports real-time inference on edge with 9x faster inference performance than the OpenPose model. 

PeopleSemSegNet, a semantic segmentation network for people detection.

A variety of computer vision pretrained models in various industry use cases, such as license plate detection and recognition, heart rate monitoring, emotion recognition, facial landmarks, and more.

CitriNet, a new speech-recognition model that is trained on various proprietary domain-specific and open-source datasets.

A new Megatron Uncased model for Question Answering, plus many other pretrained models that support speech-to-text, named-entity recognition, punctuation, and text classification.

Training support on AWS, GCP, and Azure.

Out-of-the-box deployment on NVIDIA Triton and DeepStream SDK for vision AI, and NVIDIA Jarvis for conversational AI.  ... '

Which VPN are Secure?

Useful look provided.  Who do we trust?

Which VPN Providers Really Take Privacy Seriously in 2021?

June 14, 2021 by Ernesto Van der Sar  in Torrentfreak

HOME > TECHNOLOGY > VPN PROVIDERS >

Choosing the right VPN can be a tricky endeavor. There are hundreds of VPN services out there, all promising to keep you private but some are more private than others. To help you pick the best one for your needs, we asked dozens of VPNs to detail their logging practices, how they handle torrent users, and what else they do to keep you as anonymous as possible.

private lockThe VPN industry is booming and prospective users have hundreds of options to pick from. All claim to be the best, but some are more privacy-conscious than others.

The VPN review business is also flourishing as well. Just do a random search for “best VPN service” or “VPN review” and you’ll see dozens of sites filled with recommendations and preferred picks.

We don’t want to make any recommendations. When it comes to privacy and anonymity, an outsider can’t offer any guarantees. Vulnerabilities are always lurking around the corner and even with the most secure VPN, you still have to trust the VPN company with your data.

Instead, we aim to provide an unranked overview of VPN providers, asking them questions we believe are important. Many of these questions relate to privacy and security, and the various companies answer them in their own words.

We hope that this helps users to make an informed choice. However, we stress that users themselves should always make sure that their VPN setup is secure, working correctly, and not leaking.

This year’s questions and answers are listed below. We have included all VPNs we contacted that don’t keep extensive logs or block torrent traffic on all of their servers. The order of the providers is arbitrary and doesn’t carry any value.

—

1. Do you keep (or share with third parties) ANY data that would allow you to match an IP-address and a timestamp to a current or former user of your service? If so, exactly what information do you hold/share and for how long?

2. What is the name under which your company is incorporated (+ parent companies, if applicable) and under which jurisdiction does your company operate?

3. What tools are used to monitor and mitigate abuse of your service, including limits on concurrent connections if these are enforced?

4. Do you use any external email providers (e.g. Google Apps), analytics, or support tools ( e.g Live support, Zendesk) that hold information provided by users?

5. In the event you receive a DMCA takedown notice or a non-US equivalent, how are these handled?

6. What steps would be taken in the event a court orders your company to identify an active or former user of your service? How would your company respond to a court order that requires you to log activity for a user going forward? Have these scenarios ever played out in the past?

7. Is BitTorrent and other file-sharing traffic allowed on all servers? If not, why? Do you provide port forwarding services? Are any ports blocked?

8. Which payment systems/providers do you use? Do you take any measures to ensure that payment details can’t be linked to account usage or IP-assignments?

9. What is the most secure VPN connection and encryption algorithm you would recommend to your users?

10. Do you provide tools such as “kill switches” if a connection drops and DNS/IPv6 leak protection? Do you support Dual Stack IPv4/IPv6 functionality?

11. Are any of your VPN servers hosted by third parties? If so, what measures do you take to prevent those partners from snooping on any inbound and/or outbound traffic? Do you use your own DNS servers?

12. In which countries are your servers physically located? Do you offer virtual locations?

Important note: services that offer dedicated or fixed IP-addresses are often able to link the IP-address to the user account, irrespective of the answer to question 1.

Tip: Here’s a list of all VPN providers covered here, with direct links to the answers. Some links in this article are affiliate links. This won’t cost you a penny more but it helps us to keep the lights on. ... '

Indoor Spy Drones: Always Home Cam

Not quite understanding this, except for unusual purposes, say maybe recording meetings or filming demonstrations?   There are lots of people filming things on Youtube these days, say a cooking demo.    But would it be controllable enough for that?  Seen previous examples of indoor drones, but none successful to my knowledge.  This one like others is noisy indoors, so not really a spy drone.   I like the general idea of having an indoor drone checking out the place, challenging invaders.  Taking poles from inhabitants?  Reminding them to do chores?  Make sure all are OK.   But really?  Expensive.  

Ring's New Always Home Cam Is Actually an Indoor Spy Drone

500+Gizmodo Security by Sam Rutherford 

Among an avalanche of new Echo and Ring devices announced by Amazon today, there’s one that stuck out as a clear step towards our dystopian future: the Ring Always Home Cam.

While its name might suggest that the Always Home Cam is simply just an updated home security cam, it’s way more than that. Ring describes the $250 Always Home Cam as an “autonomous indoor camera that will automatically fly to predetermined areas of the home, giving you multiple viewpoints with just one camera,” but you can’t fool me, I know a drone when I see one.

Based on Amazon’s press images, the Always Home Drone features a set of rotors on top enclosed in a plastic shroud, along with a camera built into a rectangular shaft that hangs below. When not in use, the shaft docks into the Always Home Cam’s cradle, presumably to recharge and offload any backup video recordings.

When it comes to privacy, Ring says the Always Home Cam “only records when in flight; when it’s not in use it sits in a doc and the camera is physically blocked. And, it’s loud enough so you hear when it’s in motion.” In other words, anyone who doesn’t want to be recorded will be able to hear this thing coming by listening for the signature buzz emitted by so many drones. So much for spying on the kids, but you’ll definitely be able to freak out pets.  ... '

Data and Analytics for Better Decisions

From MIT Sloan, SAS, some thoughts on decisions for analytics.  I link to items I have read and liked.  All accessible from top link.

Data and Analytics for Better Decisions

Stepping up to business challenges and opportunities means knowing how to find relevant data — and put it to work. Free access to these four MIT Sloan Management Review articles is provided courtesy of SAS   

1. To Succeed With Data Science, First Build the ‘Bridge’  

2. Demystifying Data Monetization 

3. The Recession’s Impact on Analytics and Data Science 

4. Data Science, Quarantined 

...'  

AI Battling Retail Out-Of-Stocks

I recall a related  approach suggested in the enterprise, but the right kind of data lacking,  then the right data applied.  Sounds like a fix found here?  

Can AI solve e-grocery’s erratic out-of-stock substitutions?  by Tom Ryan  in Retailwire

As out-of-stocks became increasingly common in the early months of the pandemic, Walmart’s personal shoppers turned to artificial intelligence to help find the best substitutes.

“The decision on how to substitute is complex and highly personal to each customer,” said Srini Venkatesan, EVP, Walmart Global Tech, in a blog entry. “If the wrong choice is made, it can negatively impact customer satisfaction and increase costs.”

In the past, personal shoppers would determine the best substitute themselves, but Walmart found nearly 100 different factors can go into that decision. Mr. Venkatesan added, “Trying to account for all of these would not only be too difficult, but it would also be incredibly time consuming.”

The AI technology uses hundreds of variables — including size, type, brand, price, aggregate shopper data, individual customer preference and current inventory — to determine the next best available item. It then preemptively asks the customer to approve the substituted item. Whether the substitute is approved or rejected, the information is fed back into the AI’s algorithms to improve the accuracy of future recommendations. Following the technology’s deployment at Walmart, customer acceptance of substitutions increased to over 95 percent.  ... " 

Retailwire Research on Retail Digital Transformation.

Always interesting regarding retail research activities.

Retailwire: Research    https://www.retailwire.com/  

Our recently released study, "The Beating Heart of Digital Transformation," confirms that retail planning, analytics and business intelligence apps are indeed fueling the changes rolling out at record pace in the industry. We've got two convenient ways for you to learn the key takeaways from our research.

See the on-demand webinar for a summary presentation of the results and great insights and advice from our expert panel.

Watch the recorded webinar 

Read the report ,  for a thorough look at how advanced retailers have become in their integration of these tools into their departments, their satisfaction level with the methods being using and what hurdles stand in the way of progress.

Download the report

Thanks for your interest. Please let us know your thoughts on the findings.

Regards,  Rick Moss, President, Co-founder, RetailWire

Cloud Computing Under Attack

Thoughts on vulnerability of the cloud. 

Russia’s Hacking Success Shows How Vulnerable the Cloud Is

The cloud is everywhere. It’s critical to computing. And it’s under attack.

By Bruce Schneier, a fellow and lecturer at the Harvard Kennedy School, and Trey Herr, the director of the Cyber Statecraft Initiative at the Atlantic Council’s Scowcroft Center for Strategy and Security.

U.S. Deputy Attorney General Jeffrey A. Rosen at a press conference concerning a hacking campaign tied to the Chinese government at the U.S. Department of Justice in Washington on  Sept. 16, 2020.

Russia’s Sunburst cyberespionage campaign, discovered late last year, impacted more than 100 large companies and U.S. federal agencies, including the Treasury, Energy, Justice, and Homeland Security departments. A crucial part of the Russians’ success was their ability to move through these organizations by compromising cloud and local network identity systems to then access cloud accounts and pilfer emails and files.

Hackers said by the U.S. government to have been working for the Kremlin targeted a widely used Microsoft cloud service that synchronizes user identities. The hackers stole security certificates to create their own identities, which allowed them to bypass safeguards such as multifactor authentication and gain access to Office 365 accounts, impacting thousands of users at the affected companies and government agencies.

It wasn’t the first time cloud services were the focus of a cyberattack, and it certainly won’t be the last. Cloud weaknesses were also critical in a 2019 breach at Capital One. There, an Amazon Web Services cloud vulnerability, compounded by Capital One’s own struggle to properly configure a complex cloud service, led to the disclosure of tens of millions of customer records, including credit card applications, Social Security numbers, and bank account information.

This trend of attacks on cloud services by criminals, hackers, and nation states is growing as cloud computing takes over worldwide as the default model for information technologies. Leaked data is bad enough, but disruption to the cloud, even an outage at a single provider, could quickly cost the global economy billions of dollars a day.

Cloud computing is an important source of risk both because it has quickly supplanted traditional IT and because it concentrates ownership of design choices at a very small number of companies. First, cloud is increasingly the default mode of computing for organizations, meaning ever more users and critical data from national intelligence and defense agencies ride on these technologies. Second, cloud computing services, especially those supplied by the world’s four largest providers—Amazon, Microsoft, Alibaba, and Google—concentrate key security and technology design choices inside a small number of organizations. The consequences of bad decisions or poorly made trade-offs can quickly scale to hundreds of millions of users.  ... ' 

Channeling the Inner Voice of Robots

Like consciousness or a cognitive infrastructure that feeds into our decisions and broad behavior like a consciousness?     If it produces better measure of decisions, why not.   Consciosness, inner voice 

Channeling the Inner Voice of Robots  By Samuel Greengard, Commissioned by CACM Staff, June 29, 2021

Philosophers, psychologists, and neuroscientists have long studied how and why humans talk to themselves as they navigate tasks, manage decisions, and solve problems. "Inner speech is the silent conversation that most healthy human beings have with themselves," says Alain Morin, a professor in the Department of Psychology at Mount Royal University in Alberta, Canada.

Now, as robotics marches forward and devices increasingly rely on a combination of machine learning and conventional programming to navigate the physical world, researchers are exploring ways to imbue machines with an inner voice. This "self-consciousness" could provide insight and feedback into how and why these systems make decisions, while helping them improve at various tasks.

The impact on service robots, computer speech systems, virtual assistants, and autonomous vehicles could be significant. "A cognitive architecture for inner speech may be the first step toward functional aspects of robot consciousness. It may represent the beginning of a new domain for human-robot interactions," explains Antonio Chella, professor of robotics and director of the RoboticsLab at the University of Palermo in Italy.

Robot Talk

Applying the principles of human speech and cognition to machines is a steep challenge.  "Consciousness is a very complex and fuzzy term," observes Angelo Cangelosi, professor of machine learning and robotics at the University of Manchester in the U.K. "While machines may not be aware in the way humans are aware, the idea of modeling speech characteristics to enrich interactions could deliver deeper insight into machine behavior."

The idea is taking shape. In April 2021, Chella and University of Palermo research fellow Arianna Pipitone equipped a robot named Pepper from SoftBank Robotics with a cognitive architecture that models inner speech. This allowed the robot to reason and interact at a deeper level—and to generate vocal feedback about how it arrived at answers and actions. This is possible because the parameters and attributes of the inner voice are different than those for outward expression, the researchers note.  ... ' 

What is Slam?

Below just an outline, but something I am investigating,  detail at the link.

The definitive guide to SLAM & mobile mapping

SLAM 101   ....  How does SLAM work  ....

How does it work, and what does it mean for mobile 3D mapping?

SLAM 101

Simultaneous localization and mapping (SLAM) is not a specific software application, or even one single algorithm. SLAM is a broad term for a technological process, developed in the 1980s, that enabled robots to navigate autonomously through new environments without a map.

Autonomous navigation requires locating the machine in the environment while simultaneously generating a map of that environment. It’s very difficult to accomplish, because the machine needs to have a map of the environment to estimate its own location. But to generate the map, it needs to know its own location.

As a result of this never-ending circle of dependencies, SLAM was sometimes called a “chicken or egg” problem.

How does SLAM work?

There are many approaches to SLAM. Luckily, we can still make some generalizations to demonstrate the basic idea.

Here’s a very simplified explanation: When the robot starts up, the SLAM technology fuses data from the robot’s onboard sensors, and then processes it using computer vision algorithms to “recognize” features in the surrounding environment. This enables the SLAM to build a rough map, as well as make an initial estimate of the robot's position.

When the robot moves, the SLAM takes that initial position estimate, collects new data from the system’s on-board sensors, and makes a new (and improved) position estimate. Once that new position estimate is known the map is updated in turn, which completes the cycle.

By repeating these steps continuously, the SLAM tracks the robot's path as it moves through the asset. At the same time, it builds a detailed map.  ... '

Mondelez Does Low-Code, No-Code Development

I think this is ultimately the approach we will be taking.   Not quite here yet, but it will be too risky to have individual developers do most coding work. 

Mondelez Takes a Low-Code, No-Code Approach to Development

Snack food giant Mondelez aims to get ahead of the curve in software development by building up its citizen developer community.

Joao-Pierre S. Ruth  Senior Writer  InformationWeek.

A low-code, no-code development campaign that began one year ago at snack maker Mondelez International, which owns such brands as Chips Ahoy, Oreo, Wheat Thins, and Trident, is starting to show it can scale up across the enterprise.

Sanjay Gurbuxani, global digital innovation lead for the entire enterprise and the AMEA regional CIO, says Mondelez began its journey into no-code and low-code development in spring 2020 making use of the Quickbase platform and sees its use growing throughout the organization.

Gurbuxani’s regional duties covers Asia, the Middle East, and Africa, which he says generates about a $6 billion business, or a quarter of Mondelez’s global revenue.

Wearing global and regional hats with the company, he is responsible for bringing in new and emerging technologies to disrupt the way Mondelez thinks and does business. Gurbuxani spoke to InformationWeek about why his company turned to low-code, no-code development last year with plans for widespread internal adoption that are on track to outpace market adoption.   .... " 

FBI Running Encrypted Phones

Bruce Schneier posted about the FBI effort to create an encrypted phone App.  And makes interesting points about trust and security.    Just read, worth considering

FBI/AFP-Run Encrypted Phone     Bruce Schneier

For three years, the Federal Bureau of Investigation and the Australian Federal Police owned and operated a commercial encrypted phone app, called AN0M, that was used by organized crime around the world. Of course, the police were able to read everything — I don’t even know if this qualifies as a backdoor. This week, the world’s police organizations announced 800 arrests based on text messages sent over the app. We’ve seen law enforcement take over encrypted apps before: for example, EncroChat. This operation, code-named Trojan Shield, is the first time law enforcement managed an app from the beginning.

If there is any moral to this, it’s one that all of my blog readers should already know: trust is essential to security. And the number of people you need to trust is larger than you might originally think. For an app to be secure, you need to trust the hardware, the operating system, the software, the update mechanism, the login mechanism, and on and on and on. If one of those is untrustworthy, the whole system is insecure. ...  '

It’s the same reason blockchain-based currencies are so insecure, even if the cryptography is sound.  ...'

AI Taking Over Gadgets

Now lets make them as secure as possible.    And the AI does not have to be created on the edge, it can  be encrypted and fed to devices.    We will see both.  

How AI Is Taking Over Our Gadgets   By The Wall Street Journal. July 1, 2021

 Artificial intelligence (AI) is taking over "edge" devices, including smartphones, automobiles, drones, home appliances, industrial sensors and actuators, and other devices with a microchip and some memory.

An example of this is Apple's Siri assistant, which soon will begin processing voice on iPhones using a "neural engine," rather than sending an audio recording to the cloud for processing. While AI in the cloud can be trained on the fly and improved over time, edge AI systems are pre-trained and updated periodically.

Purdue University's Elisa Bertino said one edge device could collect data and then pair with a more powerful device that can integrate data from a variety of sensors, like how a smartwatch with a heart-rate monitor transmits the data to the user's smartphone for additional data analysis.

Gartner's Eric Goodness said edge AI systems could become powerful enough to gather data and use it to train their own algorithms.  ... '

Reimplementing Software Interfaces is Fair Use

 Somewhat surprising,  and Beyond the US?   Good overview of the technical implications.

Reimplementing Software Interfaces Is Fair Use

By Pamela Samuelson  Communications of the ACM, July 2021, Vol. 64 No. 7, Pages 24-26

10.1145/3466607

A long-standing, generally accepted norm in the computing field distinguishes between software interfaces and implementations: Programmers should have to write their own implementing code, but they should be free to reimplement other developers' program interfaces. This norm, of which Sun Microsystems, the developer of Java, was once the software industry's foremost proponent, is now the law of the land in the U.S. after the Supreme Court's decision in Google Inc. v. Oracle America, Inc., which overturned a lower court ruling that reimplementing an interface infringed copyright.

The Supreme Court took Google's appeal on two issues. One was whether program interfaces are protectable by copyright law. The Supreme Court declined to decide that issue, even though many amicus curiae (friend of the court) briefs filed by software developers, organizations such as the Electronic Frontier Foundation, the Center for Democracy & Technology, and the Computer & Communications Industry Association, as well as numerous intellectual property scholars, supported Google's argument that program interfaces are uncopyrightable.

A second issue was whether Google's reimplementation of 11,500 declarations from 37 Java Application Program Interface (API) packages in its Android smartphone platform was fair use or infringement. Although a jury rendered a verdict in favor of Google's fair use defense after a two-week trial, the Court of Appeals for the Federal Circuit (CAFC) overturned this verdict. The CAFC concluded that no reasonable jury could have found Google's appropriation of that many lines of computer code was fair use. This is the ruling that Supreme Court's decision reversed.

The penultimate sentence in Justice Breyer's opinion for the 6-2 majority succinctly states the Court's conclusion: "where Google reimplemented a user interface, taking only what was needed to allow users [that is, programmers] to put their accrued talents to work in a new and transformative program, Google's copying of the Sun Java API was a fair use of that material as a matter of law."

After explaining Oracle's claims against Google, this column reviews the Court's reasons for rejecting Oracle's arguments on the fair use issue.  ... ' 

A Battery Free Internet of Things

 A look at current state and technology advances and likely futures.   We looked at these for RFID identification applications.  Implications for applications, especially for local AI applications.

A Battery-Free Internet of Things  By Esther Shein

Communications of the ACM, July 2021, Vol. 64 No. 7, Pages 16-18 10.1145/3464937

When NVIDIA purchased mobile-chip designer Arm Holdings from SoftBank last year, NVIDIA CEO Jensen Huang made the bold prediction that in the years ahead, there will be trillions of artificial intelligence (AI)-enabled Internet of Things (IoT) devices. Regardless of whether that holds true, it is safe to say the growth of IoT devices is exploding. All those devices will require power sources, and the way Josiah Hester sees it, that's problematic for the environment and society.

"When I see the 'trillion' number, I see a trillion dead batteries, basically," says Hester, an assistant professor of computer engineering at Northwestern University. "There's piles of batteries in landfills in China and elsewhere sitting there unrecycled; or they're put in furnaces and melted down, which is not a carbon-neutral event."

As a native Hawaiian, Hester also is concerned about the impact of micro-plastics and dead batteries turning up in oceans, and about lithium mining, which uses water supplies that people depend on to live. That got him thinking about how to design computer systems without batteries that instead harvest energy, thus reducing their carbon footprint and the impact on the environment.

Hester and other researchers at Northwestern designed a battery-free Nintendo Game Boy that is powered by button presses and sunlight, harvesting energy from the movement of tiny magnets and through tightly wound coils every time a user presses a button.

Now, the team is working on smart face masks that are powered by a person's breathing or movement, that will be able to capture heart or respiration rates, and also to determine whether the person is wearing the mask correctly.  ...  '

Opportunities and Dangers of Decentralizing Finance

Useful definitions and opinions on the state and future of what is being called DeFi.  Much on the Podcast and text below at the link.

The Opportunities and Dangers of Decentralizing Finance

MIC LISTEN TO THE PODCAST:

Wharton’s Kevin Werbach speaks with Wharton Business Daily on SiriusXM about the opportunities and risks of decentralizing finance.

Audio Player :Use Up/Down Arrow keys to increase or decrease volume.

Decentralized Finance — or DeFi — has experienced explosive growth in the past year. But in order for DeFi to fulfill its promise as a disintermediated ecosystem that helps rather than harms, “now is the time to evaluate its benefits and dangers,” write Wharton legal studies and business ethics professor Kevin Werbach and David Gogel, a recent Wharton MBA graduate, in the article that follows. Werbach is author of the book The Blockchain and the New Architecture of Trust and leads Wharton’s Blockchain and Digital Asset Project. Werbach and Gogel recently collaborated with the World Economic Forum to create the Decentralized Finance (DeFi) Policy-Maker Toolkit,  providing guidance to regulators and blockchain watchers everywhere.

Intermediaries have always played essential roles within financial markets, facilitating trust, liquidity, settlement, and security. Yet these benefits come with costs. Intermediation contributes to slow settlement cycles, inefficient price discovery, and limitations on market access. Financial services markets tend to be highly concentrated, with a few powerful intermediaries exercising significant control and extracting substantial rents. Since the 2008 Global Financial Crisis, there has been increased attention on structural inequalities and hidden risks of the financial system. Recent controversies such as the GameStop short squeeze, in which retail investors were blocked from trading during a period of volatility, also cast a spotlight on the shortcomings of legacy financial infrastructure.

Until now, however, intermediation was a necessary feature of finance. Even peer-to-peer fintech lending platforms such as Prosper and cryptocurrency exchanges such as Coinbase retain an important central role. This is the environment in which Decentralized Finance (DeFi) has emerged .... ' 

On Autodesk Tandem

My continued look at Autodesk's Tandem Digital Twin Solution.  Look for more posts here on digital twins.

DIGITAL TWIN FOR SMARTER OPERATIONS

Autodesk Tandem

Autodesk Tandem™ digital twin solution is here to free the data, to organize the data, and to standardize the data. The platform enables project teams to deliver an effortless handover, resulting in smarter operations for owners.

What is Autodesk Tandem?

This platform harnesses BIM data to create a digital twin of a facility. At the end of a project, your team can deliver to owners a comprehensive digital handover of easily accessible and insightful data that makes for ready-to-go operations.

Diagram of building phases: plan, design, build and operate.

How does Autodesk Tandem work?

It federates the data created during a project into one comprehensive digital replica. From planning to operations, it enables project teams to specify and capture data throughout a project’s lifecycle, so they can deliver a digital handover to the owner for better facility operations.

How does Autodesk Tandem help your project teams?

The digital solution empowers AEC firms to harness data. About 80% of the lifetime cost of a facility is in operations. By harnessing all the data created during a project, you deliver more value to customers. Autodesk Tandem enables more efficient operations, reduces waste, and lowers maintenance costs.

What goes into creating a digital twin handover?

Autodesk Tandem allows for a more integrated workflow involving the following stages:

1. Specify  Set up, track, manage, and define the data requirements most relevant to your project and organize them to best suit your needs.

2. Capture  Collect and aggregate the data required for your digital handover from all involved project teams and contributors into one digital hub.

3. Verify Validate that all the data is complete and accurate for an effortless digital handover.  ... '

Magnetic Memory System for AI

Seems a memory device meant for learning from data  in AI system.  Faster it implies.

A More Robust Memory Device for AI Systems

By Northwestern University McCormick School of Engineering

A research team from Northwestern Engineering and the University of Messina in Italy have developed a magnetic memory device that could lead to faster, more robust artificial intelligence systems.

Artificial intelligence systems could be strengthened by a new magnetic memory device designed by researchers at the Northwestern University McCormick School of Engineering (Northwestern Engineering) and Italy's University of Messina.

The device is fabricated from antiferromagnetic (AFM) materials, which boast innately faster dynamics than ferromagnetic materials, and cause no unwanted magnetic interactions; AFM-based memory cannot be deleted with external magnetic fields.  The researchers used the manufacturing-conducive iridium manganese AFM system to build a memory device that expands on an earlier silicon-compatible device.  The new device can write data, and more reliably read out information once it has been written.

Said Northwestern Engineering's Pedram Khalili, "Our technology is general-purpose and could be applied anywhere memory is used in high-performance computing systems today."

Home Networking by Radar

Interesting.   Security?  Might this kind location could be combined with other kind of location and movement information?  

Home networking by radar

OmniConnect project: Radar beams for networking and localizing everyday objects

Research News / July 01, 2021

In the OmniConnect project, Fraunhofer researchers are working with other partners on networking objects in indoor areas. They are doing this using radar beams and passive tags that are attached to moving objects, but also to people. This technology effectively detects the positions of the tags and therefore of the objects as well. It can also be used in the care sector, to avert dangers to people who are prone to falling.

For many of us, the term “radar” conjures up scenarios from the aviation or shipping industries. Radar technology is also deployed to search for space debris. This always involves localizing and measuring the velocity of flying objects. In recent years, however, scientists have been seeking to scale up the use of radar beams for applications inside closed rooms. The Fraunhofer Institute for Reliability and Microintegration IZM in Berlin is focusing on a most promising project in this area.

In the OmniConnect project, a radar is used to detect the motion and position of objects inside rooms. The researchers are using what we refer to as a secondary radar. A conventional radar detects objects and their movements, but does not provide any other data. A secondary radar combines radar beams with tags that are attached to objects. These passive tags not only reveal to the system position and movements inside a room, they can transmit information about the object as well.

Energy-efficient, compact, harmless

Thanks to the high frequency in the 60 GHz band, the systems can be developed for a high degree of integration. Each send and receive module is just 25 square centimeters in size. Conflicts with mobile radio networks, Wi-Fi or Bluetooth are impossible. The radiation this technology emits is completely harmless to human beings.

The system developed in OmniConnect is the ideal solution for networking any objects or everyday items with one another or integrating them into a home network. Because the passive tags do not need a separate power supply, there’s no inconvenience of having to replace batteries. ... ' 

Next Wave of Computing Innovation will be Driven by Crypto

 Andreessen Says it will all be about Crypto

Andreessen Horowitz  A16z

Crypto Fund III  by Chris Dixon, Katie Haun, and Ali Yahya

cryptocurrencies & blockchains  a16z crypto  announcements

We believe that the next wave of computing innovation will be driven by crypto. We are radically optimistic about crypto’s potential to restore trust and enable new kinds of governance where communities collectively make important decisions about how networks evolve, what behaviors are permitted, and how economic benefits are distributed. That’s why today we’re pleased to announce a new $2.2 billion fund to continue investing in crypto networks and the founders and teams building in this space. This represents the beginning of an exciting new chapter for the a16z crypto team.

This fund allows us to find the next generation of visionary crypto founders, and invest in the most exciting areas of crypto. We invest in all stages, from early seed-stage projects to fully developed later-stage networks.

Expanding our team to provide unrivaled regulatory and operational capabilities alongside our crypto-native data science and research services.

When we started a16z crypto, our ambition was to bring together a highly specialized and crypto-native team that had never existed in venture capital to support the companies and projects we invest in, help them navigate the complexities of this space, and help projects design crypto protocols and mechanisms. We’ll continue to deepen our focus on the kind of behaviors our entrepreneurs expect from a crypto investor including staking, delegating, running nodes, actively participating in governance and designing mechanisms. We’re expanding our data science and research teams to support this effort. 

As with any new computing movement, crypto has endured a variety of challenges and misconceptions. That’s why we are also bringing together heavy-hitters across several functions to help translate crypto to the mainstream. Our team now includes experts in marketing, public relations, policy, regulatory affairs, recruiting, as well as general startup management. We are promoting Anthony Albanese, who joined us last year from the New York Stock Exchange, to Chief Operating Officer to lead operations.  ....... ' 

Turing Lecture on Deep Learning

 Quite good, relatively non technical.   Worth a look.

Turing Lecture

Deep Learning for AI

By Yoshua Bengio, Yann Lecun, Geoffrey Hinton    from ACM

Communications of the ACM, July 2021, Vol. 64 No. 7, Pages 58-65   10.1145/3448250

Yoshua Bengio, Yann LeCun, and Geoffrey Hinton are recipients of the 2018 ACM A.M. Turing Award for breakthroughs that have made deep neural networks a critical component of computing.

Research on artificial neural networks was motivated by the observation that human intelligence emerges from highly parallel networks of relatively simple, non-linear neurons that learn by adjusting the strengths of their connections. This observation leads to a central computational question: How is it possible for networks of this general kind to learn the complicated internal representations that are required for difficult tasks such as recognizing objects or understanding language? Deep learning seeks to answer this question by using many layers of activity vectors as representations and learning the connection strengths that give rise to these vectors by following the stochastic gradient of an objective function that measures how well the network is performing. It is very surprising that such a conceptually simple approach has proved to be so effective when applied to large training sets using huge amounts of computation and it appears that a key ingredient is depth: shallow networks simply do not work as well.

 We reviewed the basic concepts and some of the breakthrough achievements of deep learning several years ago.  Here we briefly describe the origins of deep learning, describe a few of the more recent advances, and discuss some of the future challenges. These challenges include learning with little or no external supervision, coping with test examples that come from a different distribution than the training examples, and using the deep learning approach for tasks that humans solve by using a deliberate sequence of steps which we attend to consciously—tasks that Kahneman56 calls system 2 tasks as opposed to system 1 tasks like object recognition or immediate natural language understanding, which generally feel effortless.   ... " 

Matterport for Creating 3D Virtual Tours

Had mentioned I was looking at simplified Lidar Solutions.  Here just mentioner to me:  Matterport. Below shows some of the technology supported.  Examining this further.

3D virtual tours can take your business to a new dimension

Whether you want to give buyers the feeling of being in a new home, help guests start picturing their stay, or showcase the wonders of your venue to event planners and patrons - Matterport 3D virtual tours can help you do more business, faster.

Engage your customers like never before

People are shown to be 300% more engaged with a Matterport 3D virtual tour than they are with 2D imagery. 3D virtual tours can help you create better experiences, happier customers, and more revenue. What's more, you can start today for free with no previous photography experience by opening a free account and using your iPhone (or iPad) to capture your first space.  Get a FREE demo now and see it for yourself!   .... 

As well as creating incredible, accurate 3D virtual tours of buildings and spaces, Matterport's software is packed with great features to help you create amazing additional assets and share your tours with the world.

Share easily on social media sites, such as Facebook, Instagram, and YouTube

Publish to Google Street View, VRBO, and Realtor.com

Create detailed floor plans that give your customers the bird's eye view

Automatically generate guided tours to walk your customers through the highlights

Create 4k print-quality photos [with the Pro2 Camera]

Measure your space

Add Mattertags™  .... 

Inhalio: Digital Scent as a Service

Recently received an update of this.  Previously had examined for retail scent oriented applications as part of our innovation spaces.  Impressive new directions, like use in automobiles.   Note use for things like mood mapping.   Contact Keith for more information. 

Inhalio Overview

Inhalio is a global leader in cloud-based digital scent technology and digital scent as a service. The Inhalio Digital Scent 3.0 Platform is transformative to personal wellness and well-being by diffusing scents that sanitize airborne viruses, eliminate malodors, and create mood-mapped experiences. We focus on the science of scent using a data-centric process of precise scent infusion, molecular dry-air diffusion, and detailed customer insights. Inhalio recently acquired Exhalia, a recognized leader in scent infusion systems.

INHALIÓ: THE WORLD’S 1ST CONNECTED PLATFORM FOR SCENT ™

Inhalio pioneered the concept of a Wellness Ride™ – a breakthrough for mobility companies looking to give their customers healthier, and new emotive in-vehicle experiences. The Inhalio digital scent platform and our wellness scent services are transforming the rideshare, mass transit, and OEM marketplace.

— Keith Kelsen, CEO Inhalio

Drones for Space Rocks

 Many rocks fall, but relatively few are found.   A solution? 

Meteorite-Hunting Drones Could Help Find Freshly Fallen Space Rocks By New Scientist

A team of scientists used drones and machine learning (ML) to try to find just-landed meteorites, as part of a study funded by the U.S. National Aeronautics and Space Administration (NASA).

Researchers from the University of California, Davis (UCD) flew a camera-outfitted consumer drone over a dry lake bed in Nevada, where meteorites may have hit following a meteor fireball in 2019. ML software studied aerial images for objects resembling dark-colored space rocks, as well as for intentionally placed meteorite specimens.

The flagged specimens turned out to be terrestrial rocks.

UCD's Robert Citron said, "If we can improve our ability to search more of these small falls, then we can gain a lot more data connecting meteorite samples to their tracked incoming trajectories."

From New Scientist

Quantum Safe Trust for Vehicles

 Threats of Quantum Powered Cyberattacks, safety of vehicles and their data. 

A 2019 claim by Google AI on groundbreaking quantum computations, while welcomed by many, was met with an intensified sense of urgency by security professionals who face the challenge of how to respond to threats of quantum-powered cyber attacks.

In the Queue Case Study, "Quantum-safe Trust for Vehicles: The Race is Already On,"  Michael Gardiner, Alexander Truskovsky, George Neville-Neil, and Atefeh Mashatan discuss the potential risks posed to the automotive industry by quantum computing.

Among the concerns are over-the-air software updates for smart cars; the telemetry data cars send to manufacturers; sensors in autonomous vehicles that talk to engine control units; and the challenges the automotive industry faces in implementing quantum-resistant public key infrastructure.

Auto manufacturers can begin addressing the challenge by ensuring that engine hardware designed and built today is capable of handling the cryptography that will become essential once attackers are able to take advantage of quantum-compute capabilities. Additionally, quantum-safe algorithms can be tested on vehicle components and quantum-safe trust anchors can be embedded in vehicles.

Queue is ACM's magazine for practicing software engineers. Written by engineers for engineers, Queue focuses on the technical problems and challenges that loom ahead, helping readers to sharpen their own thinking and pursue innovative solutions.  .. '

Linking Machine Learning and Quantum Computing

Experimenting with Quantum computers using Machine Learning

 IBM Releases Qiskit Modules That Use Quantum Computers to Improve ML

VentureBeat, Chris O'Brien, April 9, 2021

IBM has released the Qiskit Machine Learning suite of application modules as part of its effort to encourage developers to experiment with quantum computers. The company’s Qiskit Applications Team said the modules promise to help optimize machine learning (ML) by tapping quantum systems for certain process components. The team said, "Quantum machine learning (QML) proposes new types of models that leverage quantum computers' unique capabilities to, for example, work in exponentially higher-dimensional feature spaces to improve the accuracy of models." IBM expects quantum computers to gain market momentum by performing specific tasks that are offloaded from classic computers to a quantum platform. ... ' 

Fast Food too Slow in Opening Dine-in?

 An example of semi-permanent change from COVID?   We certainly are doing far more drive through than before the pandemic.  Been discussing the dynamics for some time. 

Is fast food going too slow in reopening for dine-in business?  in Retailwire  by Matthew Stern

City by city and state by state, U.S. retailers are reopening as reported COVID-19 infections continue to drop. Fast food chains, however, have been moving more slowly.

While local restaurants have been opening up, chains including Popeye’s, McDonald’s, Starbucks and those owned by Yum! Brands have remained more focused on digital sales and limited in the full reopening of stores in key markets, according to Yahoo! Finance. Brands cite operating in accordance with CDC guidelines and local regulations, but are farther from across-the-board reopening than smaller restaurant operators.

Part of the hesitation may be found in the franchise model some of these brands operate under. McDonald’s franchisees in late May were insisting on staying closed despite the easing of local restrictions, according to a Bloomberg report. Some stated that they were doing well enough through drive-thrus and mobile ordering and had neither enough staff to reopen nor interested diners to justify reopening. The chain hopes to have all restaurants fully open by late August.  ... " 

Copilot Assistant Coding

 Actually did this in the early days, with human co-pilots, but it never took off.  Now with secure code more of a necessity, could be useful.  This could drive closer to Lo-Code too.   How well can this work today?  Even flagging security dangers in patterns of code could be useful.  But how well can we recognize such patterns?  Following. 

OpenAI and GitHub Unveil New Copilot AI Assistant for Coding

ERIC HAL SCHWARTZ in Voicebot.AI

A new virtual assistant created by OpenAI and GitHub will suggest code to software developers as they work. The new GitHub Copilot tool leverages an improved version of OpenAI’s popular GPT-3 language model called Codex to teach the AI how to collaborate in a coding project like a human partner.

AI COPILOT

Github Copilot takes the concept of natural language processing and applies it to programming languages. The idea is to imitate a “pair programmer,” when two developers simultaneously work on a coding project and comment and annotate each other’s work along the way. The AI theoretically takes the junior partner role in the endeavor, making its name entirely apropos. Copilot relies on OpenAI’s Codex model to understand what the programmer is doing and come up with suggestions. Like GPT-3, Codex is built on an enormous collection of data to teach an AI how to suggest a line or more of code. The AI learns from what suggestions the human user accepts or rejects, honing its understanding and ideally leading to better code ideas.  ... ' 

NASA Talent Mapping

 Took a number of looks at advanced 'talent mapping' in the enterprise.  Now is the time this could be very useful especially if we can measure talent well, predict needs and deliver.

NASA Is Using Data Science to Fill Its Data Science Skills Gap,  By ZDNet, June 30, 2021

NASA is creating a workforce talent-mapping database to identify the data skills required for all kinds of projects within the space agency.

"You know, we're NASA, so we're doing a lot of that type of stuff," says David Meza, acting branch chief of people analytics and senior data scientist. The goal is to identify capability that already exists within the organization.

Fittingly then, the solution to filling the data science skills gap at NASA lies in data science itself. The database that Meza's tean is developing uses Neo4j technology to build a knowledge graph, which is designed to show the complex and varied relationships between people, skills, and projects at NASA.

The team initially focused on creating an occupational taxonomy, which analyzed the various components of a role from an employee, training, and project perspective. The team built a model and to start identifying people with skills in specific occupations. That model highlights the kinds of abilities that other individuals in NASA might need to complete tasks in each occupation successfully.

Flying Cars Here Yet?

Depends how you define them.   Right now the definition means they can be driven around like a car and then take to the air as needed.   Such things have existed for some time, but not yet even close to mass-produced and generally available.   Formal  licensed pilots are still involved.  And here no provision for vertical take off, so needs an airport.   Autonomy/driverless is sometimes also mentioned, but not usually carried in the general definition.

Flying Car Makes 35-Minute Test Flight Between Cities   By Ryan Whitwam in ExtremeTech

Whenever the subject of the future comes up, it’s always the same: Where are the flying cars? Well, you could argue that flying cars are finally here. A hybrid aircraft known as the AirCar has just completed its first successful flight between two airports. It’s still a prototype, but it only looks a little bit ridiculous. 

The AirCar’s designer, Stefan Klein, piloted the vehicle for an uneventful 35-minute flight between international airports in Nitra and Bratislava, Slovakia. He said the flight was “normal” and “very pleasant.” The AirCar has now passed 40 hours in the air, which is an impressive feat considering how many “flying cars” have never gotten off the ground. 

Unlike many past attempts at a hybrid car-aircraft, the AirCar is not equipped for vertical takeoff and landing. That makes it less practical for the average driver with aspirations of taking to the skies because you need to find a runway. The car also needs some time to get ready for flight. It takes about 15 minutes for the motors inside the body to extend the wings and tail. You’ll have to deviate the same amount of time to stowing the wings when you land. At that point, the AirCar isn’t probably going to cut down on your morning commute.  ... "